TW490981B - Method and device for code group identification and frame boundary synchronization - Google Patents

Abstract

The invention provides a method for code group identification and frame boundary synchronization, which includes the following steps: (a) provide a plurality of the secondary synchronization code sequences (SSCSs) with length of L codes, in which each SSCS corresponds to a code group; (b) sensing and sampling the transmitted signals of each base station to form a reference sequence, which includes the sampled count less than or equal to L codes; (c) comparing the reference sequence with the secondary synchronization code sequence to find one or more most similar sequence; (d) when the most similar sequences are not unique, increasing the sampled count codes of the reference sequence and going back to step (c); and (e) when the most similar sequence is unique, outputting the related code group of the most similar sequence and a frame boundary.

Description

490981 V. Description of the invention (1) The invention relates to a method and device for code group identification and frame boundary synchronization, and a base station (ceU) search method, especially a second-generation mobile communication cooperation plan (3rd generati ο η

A method for searching a base station of a wide division (multiple-band) code division multiple access (CDMA) system proposed by the Partnership Project ′ 3GPP).

At present, direct-sequence CDMA mobile phone systems (cel lular systems) can be roughly classified into two types. One is a synchronization system that requires precise synchronization between base stations (inter-cel 1), and the other is an asynchronous system that does not require synchronization between base stations. For the synchronization system, the hand-of f of the mobile phone between base stations will be very fast, because the starting time of the identity code of the neighboring base station (identica 1 c ode) is only the base station that is currently connected. The stage exhibits different fixed offsets. However, in order to achieve the purpose of synchronization, each base station also needs a lot of expensive equipment, such as a global positioning system (GPS) and a precise timing oscillator. The timing oscillator provides a base clock, and G P S provides the basis for clock synchronization. In addition to the expensive equipment, for a base station that has difficulty receiving GPS signals, such as

It is said that in the basement or dense residential area, the implementation of the synchronization system is also difficult.

As for the asynchronous system, such as the broadband CDMA system proposed by 3GPP in Europe, each base station uses two (synchronization channels), and the synchronous channels transmitted to the two synchronous channels are shown in FIG. 1. By getting the code ’on the mobile (e.g. mobile

490981 V. Description of invention (2) Telephone) can establish a good link and will not cause disconnection when the base station is switched. These two synchronization channels are the primary synchronization channel (PSCH) and the secondary synchronization channel (SSCH). 0 In an asynchronous system, the PSCH is a channel common to all base stations. (Slot) The primary synchronization code (PSC) sent once, marked as Cpsc, consists of

The length of one PSC is 256 code bits (chip). And because the period of the psc is a 0η * slot ', sending out the PSC will also send the time slot synchronization signal of the downlink link channel, as shown in Figure 1. SSCH is an identification sequence composed of 15 secondary synchronizati = de (SSC), which are respectively labeled as i in code group and k in sequence. The identification sequence is sent continuously with a time frame as a cycle. Each ssc is composed of 16

Orthogonal codes are selected from the group, and the length of each 4 ortho code is 2 5 6 code bits. The ssc in the recognition sequence is one by one and the PSC is tied and sent out by the base station at the same time. The identification sequence belongs to one of different groups: the code group, which is also the basis of a base station's downlink $, = downlink scrambling code. In ancient times, it consists of a sequence of 15 orthogonal codes. &Amp; Sentences: ΐ :: Sequence position is shown in Figure 2. 6 The 4 specific series are the series produced by the 仏 shifts of ~ are all unique. ^ 96〇 (-64 * 15) are different from each other. Use this

Page 5

Find out the code group used by a base station and the starting position of the% box. When the mobile terminal of the broadband CDMA system proposed by 3jPP conducts a base station search: find Π: ί first find a base station that can provide the largest signal, and then ™ the scrambled code and time frame of the downlink link of a found base station . : ': The base station search can be performed through the following three steps: One step ... The time slot synchronization amount is ρ ^ Γ 3. When the base station is the base station, the mobile terminal will first find the signal that can reach the PSCH at 1 ^ 11. Secret ^ and its timing. This action is usually completed by a corresponding state. It is precisely because each base station has the largest J-signal energy in the PSCH, which corresponds to the ability to improve. And by the timing of the psc You can determine the maximum number: the slot length of the base station and the slot boundary. Because of &, the mobile terminal :: the entire internal clock (clock) and the 0 · ^ slot that provides the largest signal Synchronization Step 2: After the code group identification is synchronized with the frame boundary, the time slot can be used to find out when the first step belongs to the 64 code groups. When can you start? Frame into a sequence, and then the number of sequences that can be produced by shifting the sub-Semicircle is unique. The mobile terminal can collect the SSC in the SSCH, which code group the identified sequence sent by the base station, and A time frame should be the frame boundary synchronization. An intuitive implementation of the SSC in 15 consecutive time slots in the length of the received sequence and 9 6 0 sequence from 6 4 code groups and 1 5 In comparison, because this 9㈤, you can Find out in exactly the same number

490981 V. Description of invention (4) column. In this way, we can know which code group group the base station found is sending. It can also know how many timeslots should be cyclically shifted to be the start time of the time frame of the found base station to achieve the purpose of time frame synchronization. Step 3: Identification of scrambled codes After completing the above two steps through two synchronized channels, the mobile terminal then uses a common pilot channel (CPICH) to change all the codes corresponding to the code group found in step 2. Main scramble code-check 'to identify the pSC currently used by the base station. Then, the primary common control physical channel (PCCPCH) can be found. Then, the system and base station information can be read. In summary, the main objectives of searching for a base station can be divided into (丨) a base station that detects the maximum signal energy, (2) acquiring an identification code group (code gr〇up) synchronized with the frame boundary, and (3) acquiring a downlink Link master scrambled code. As mentioned earlier, the most intuitive way to find the base station's code group and time frame boundaries is to compare the received ssc sequence with the possible 960 sequence. In this way, the code group of the base station can be found at the same time with a 4 frame boundary. However, if such a method is to be performed, there may be two memory or registers for storing 960 SSC sequences, and a device for comparing two SSC sequences with = is costly. Another, and compared, the switching actions required for the SSC sequence will cause a large number of aspects. J: This is very bad for the mobile end in mobile communication. The main purpose of the present invention

Good code group identification

Is to provide a

490981 V. Description of the invention (5) Method and device for synchronizing with frame boundary. The method and device of the present invention have fault tolerance and save power. In order to achieve the purpose of the present invention, the present invention provides a method for code group identification and frame boundary synchronization, which includes the following steps: (a) providing a secondary length of L codes (c) Synchronization code sequences (SSCSs) 'Each SSCS corresponds to a code group; (b) sensing and sampling signals transmitted by a base station to form a reference code sequence (reference code sequence, RCS), the reference sequence contains codes with less than L samples; (c) comparing the reference sequence with the secondary synchronization code sequence to find one or more most similar sequences; (d) When the most similar sequence is not unique, increase the number of samples of the code in the reference sequence and return to step (c); and (e) when the most similar sequence is unique, output the corresponding to the most similar sequence Code group and a frame boundary ° On the other hand, the present invention provides a device for identifying a code group and synchronizing with a frame boundary. The device of the present invention includes: a memory, a recording device, a plurality of processors, and a control device. The memory ′ is used to store a plurality of code groups (c o d e g r 〇 u p s), and each code group includes a specific sequence (predetermined code sequence). The recording device is used to record a signal transmitted by a base station to form an incoming sequence; each processor processes a corresponding code group, and each processor includes a first sampling device and a second sampling device. Device and a search device. The first sampling device is used for inputting a part of the input

Page 8 ^ 0981 ^ 0981 V. Description of the invention (6) The sequence is set to the complex reference sequence that can be generated. The test sequence is the most similar one. The sequence is only the code group and the control device. The new triggering of these sequences is the only one in this invention. This particular test sequence can make this 'required less' naturally important. Similar sequence numbers are required. For the purpose of this article, we will give a special example: to form a first reference sequence. The specific number sequence corresponding to the corresponding code group of the second sampling device portion is sampled separately to form a plurality of second search devices, and the first reference sequence and the second parameters are used Compare to search for one or more single-code groups. The control device finds one or a plurality of the most similar series among the most similar of the single code group. When the most similar one, the control device outputs a frame boundary corresponding to the most similar series. When the most similar series is not unique, the first and second sampling devices increase the number of samples, and the processing device is weighted. The advantage until the control device finds the most similar is that the length of the reference sequence must not be the same as the length of each sequence, but only a smaller length parameter is needed to perform the code group identification and frame boundary synchronization. . Because the register and related switch operation will be relatively reduced, it can save power. Power saving is a very important element in mobile communications. At the same time, the present invention is to find the most, so even if the letter f with a little error in the reference sequence has a correct code group and pivot phase. ^ 1 ^ The above objectives, features and advantages can be more clearly understood. 乂 4 土 贝 例 例 'and in conjunction with the attached drawings, detailed description such as

490981 V. Description of the invention (7) Fig. 1 is a structural diagram of two synchronization channels; Figs. 2A and 2B are an arrangement sequence of scc sequences of 64 code groups. Fig. 3 is a diagram of a device using the present invention. Architecture diagram; Figure 4 is a flowchart of the method of the present invention; Figure 5 is a schematic diagram of the comparison between the reference sequence and the secondary code sequence of the code group; ^ Figure 6 is the search for the most similar A series of flow-based intents; Figure 7 is a schematic diagram of the implementation of the present invention with a DSP chip; Figure 8 is a sliding window ratio_intention in Figure 5; a solid comparison is a schematic diagram of a possible logic circuit ; And FIG. 9A and FIG. 9B are schematic diagrams of the operation of the controller, respectively. The description of the two window size control symbols in Figure 8: 3 1 ~ Correlator; 34 ~ Maximum Selector 3 8 ~ Searcher; n73 ~ Lookup Table, · Implementation of 81 82 88-92 Example N 83, 85 ~ Count '94 84, 86, 8b, window size controller; 87 ~ it 3 2 ~ similarity store 36 ~ recorder; 71 FIFO memory; shift register module; bit controller 8 a ~ Adder; the spirit of the total line is to use-a small section from the received reference sequence to page 0. 5. Description of the invention (8) Ϊ The second graph / all code groups are compared to find the most similar It can be obtained that the code group and frame phase used by the currently connected base station produce Alt: 2. You can know the 64 SSC sequences represented by the 64 code groups (length == generated by cyclic shift). The sequence is only-, which is also the main focus of == when the sequence was originally designed. However, if you are more ^ m ^ ^ ^ Mao Jian /, the SSC sequence with a length greater than 3 is present in the entire mouth / now ，, also = One. For example, (1, 1, 2, 8) is ^. So, the first four codes in the group of brothers did not appear elsewhere. ^ = If only the sequence of (1,1,2,8) is received, the code group of the base station of Λ is the first code group, and the frame phase (frame boundary) is #]. ^ ^ ^ ^ ^ ^ ^ In order to identify the code group and frame synchronization and ^ ΓΙ time all codes, as long as the fragment is enough. And Ben Moming is using this feature to implement. Please refer to Figure 3 Figure 3 is an architecture diagram of the Panmai device using the method of the present invention. Because in 3GPP, each -ssc is selected by a set of "orthogonal (〇rth〇g〇nal) codes, so" 3HH-3116 correlator to identify the current input signal)), "The order of the received ssc is marked with 'Indicate the in-phase', / Indicate the quadrature phase (gUadrature-phase), The degree of intersection code. Each correlator 3101-3116 is followed by a similarity store (Collelatlon bank) 320 1 -32 1 6 to store the degree of similarity of b in a time frame 'and The frame is a cycle, and the accumulation is repeated to reduce the effect of noise. Take the large selector 34 to find the largest of the 16 similarities and

Page 11 490981

Rotate out a number of M16 digital codes. Α determines which of the 16 orthogonal codes the current signal correction and q (m) should be. The "recorder" records ^ days and slots. The _ digit code output by the maximum selector 34 is formed into an incoming sequence. The searcher 38 then decides based on the result of the input sequence Code group and time frame phase. Please refer to FIG. 4. FIG. 4 is a flowchart of the method of the present invention. First, the method of the present invention first prepares an input sequence 40 of N codes. Next, input The first N — κ codes in the sequence are regarded as a reference sequence 42, where K is an integer greater than 0. At the same time, the 'correlator 3 1 0 1-3 11 6 continues to receive and record the received signal 4 4. Then 'compare the reference sequence with the secondary synchronization code sequence of the 64 code groups one by one to find one or more of the 64 secondary synchronization code sequences with the most similar sequence of 46. In case the most similar sequence is not unique (Yes in 4 8) 'Increase the number of samples in the reference sequence by 50 (increase p codes), p is an integer greater than 1, and is in phase with the most similar sequence 5 2 Compare to find and replace the most similar series. The circle will continue until the most similar sequence is unique. Of course, in order to avoid the unique most similar sequence that has not been found due to the small signal-to-noise ratio in the harsh environment, we have added A limit condition for the number of times is added to the loop of the number of samples of the code. If the number of loops exceeds a certain number of times A, A is an integer, (Yes in 54), the current reference sequence is discarded and the input sequence is prepared 4 0, another search is performed. When the most similar sequence is unique (N 0 in 4 8) ', it is repeated to find the unique most similar sequence (No in 56) once. After 1 time After searching (Yes in 56), 1 is at least 1,490981. V. Inventory (10) Integer, which produced the most majority table of I candidates. Also Ning ~ K where ', ,, and the number 5 8, and output the last number @ 丨 &# _ code group and a box boundary. Take the group of bulls related to the silly magic sequence. Figure 5 is the reference sequence of the present invention and the first mabin, I ίί Ϊ, Schematic diagram of comparison of Γ Shima series. To understand We have prepared a sliding window for the parent code group to find the most probable position. The size of the sliding window is the same as 2 of the reference number ^. The sliding window can be determined by the The first time slot slides to the 15th time slot. In this way, the SSC framed by the sliding window can generate ^ different sampling sequences. Comparing each sampling sequence with the reference sequence ^ can produce a comparison result V (i, X), where i represents the group of codes, and the flat number 'X represents the position of the sliding window. For example, if the ^ degree of the reference sequence 疋 6 codes' is {1 0, 1 5, 9, 10, 1 0, 2}. Because the sliding sequence generated by the first frame in the first code group is (1'1, 2, 8, 9, 10) 'The reference sequence is completely different from the code in the sample sequence, so V (1 ,, 1) is equal to 0. The sampling sequence generated by the sliding window in the second frame of the first code group is (丨, 2, 8, 9, 1〇, i 5). The reference sequence is the same as the sample sequence, so V (1 , 2) is equal to 1, and so on. That is to say, each code group will produce 15 comparison results. The largest comparison result, for example, V (1, 6), the corresponding sampling sequence becomes one of the candidates of the most similar sequence. Similarly, 64 code groups yield 64 candidates. Among them, the largest comparison result, V (k, m), corresponds to the sample sequence that is the most similar. In this way, it can be known that the code group is k 'and the frame phase starts at the m-th code.

Page 13 490981 V. Description of the invention (11) Figure 6 shows the process of finding the most similar sequence of the present invention. The 64 code groups all use the same search μ 囷 口 as an example to explain the search of the hunzeng, =, : 夂 We have a 1-inch yard of tru's beating process. When 掬 is in progress, it will perform the following steps in order: Tian Ji + 々 异 法 1 · Set m (+ ^. 2 sliding window sliding one by one to take _ two moving-number of people, brother m sampling sequence) = 1. When the number of samples in the specific car sequence is the same as the N_ one code in the McCaw sequence, the comparison result v (i, m) is increased by one. Mutian 3. If all 15 comparison results are obtained, perform the following steps: m value is increased by-, that is, the sliding window is moved down «Back to step 2. …, Silly 4. Among the 15 comparison results, select the largest comparison result, and use the corresponding reference sequence as the candidate for the most similar sequence representing code group i. 5. Among the candidates generated by the 64 code groups, select the largest comparison result V (k, m), and use the corresponding reference sequence as the most similar sequence. 6. End this algorithm. The algorithm used in the present invention can be implemented by digital signal processing (d 丨 g 丨 t a 丄 signal processing, DSP) chip or logic circuit technology. FIG. 7 is a schematic diagram of implementing the present invention with a DSP chip. The received SSC sequence first exists in a first-in-first-out (first & first out, FI F0) memory 71, and the ssc sequence in 64 code groups exists in a look-up table (LUT) In 73, when the system starts, the search algorithm, the ssc sequence in the 64 code groups, and the received SSC sequence will be downloaded into a DSP chip 72 at the same time. As previously

Page 14 490981 V. Description of the invention (12) The search algorithm basically contains 64 similar subroutines. Each subroutine is used to find a most similar sequence in a code group. In the DSP chip, all 64 subroutines are processed in parallel. Finally, the DSP chip can identify code groups and frame boundaries. Figure 8 is a schematic diagram of a possible logic circuit compared with the sliding window in Figure 5. From a lookup table, the ssc sequence of code group i is first loaded with a shift register module 86. When the maximum selector 34 outputs N digital codes to form an input sequence and record it in a shift register module 8, some of the input sequence and the SSC sequence of code group i will be controlled by the window size. The device (window size controller) 82, 85 is sent to the shift register only two groups 83 and 84. The window size controllers 82 and 85 are used to control the number of SSC entering the shift register module 83 and 84 respectively, which is / J, 〇 for sliding windows, each shift register module (81, 83, 84, and 86) are composed of SN AB day shift registers. T, SN represents a time; #, said number, Be is the minimum number of bits used to represent an SS (:. For example, SN is equal to 15, and because there are 16 (24) SCC, so compare to 8 ;. Two shifts each register 1 bit temporary storage ... and 8 = code (composed of his bits) comparison. When the shift code is the same; Ma and the shift register module 84 are rotated out The next summation line ... (round out 1. Otherwise, the comparator 87 outputs Q. And to calculate the same code ^ 目 88 and dUmP Circuit) h uses the number, which is the similar value (likelihood V. Invention Explanation (13) value) 〇 Here, the counter S8 can be used to complete the operation. Counter (counter): used to slide the viewing window module 84 has compared the number of codes. The SN codes in the register level 84 are reset by a bit register mode signal. The number of I-shots is 88. It will be rotated out—a pulse that says that the register module has been shifted. 86cyclic # 's description code. The pulse signal triggers the total line 8 ^^ temporary storage modules 83 at the same time. The condition of the handle At controller will not change. The next window The small control register module 84, which means that the codes in the # 86,: f 3 groups of 86 will be moved into the shift line and Hb f + will be moved to the next code. Then the comparison of another — = will be performed in order to Another similar value is generated. Sliding the mother s sliding window once will produce a similar value. The V (1), V (2), and 彳 t stored in the 8th shift register module 8b. From the similarity values calculated by sliding once, twice and SN times, a maximum selection circuit can be used to find a maximum value. Count the number of cycles of ^ w shift register module 86. When the meter " ^ 89 counts out the shift register module 86 returns to the original state (the mobile 7SN counter 89 will send out-pulse resets itself, on the other hand triggers the selection circuit 8c to output the most similar value found And its corresponding position 0 Figure 9A and Figure ⑽ are the window size control in Figure 8 "82 is not intended. As shown in Figure 8, the window size control

490981 V. Description of the invention (14) '85 and 82 are controlled by a window size control signal (window size control signal slg ^ nal, WSCS). Each window size controller is composed of SN * Be bit controllers (92 or 94). The bit controller is used to control whether the signal to be transmitted is an input signal from the shift register module (86 or 81) or a preset value. The preset values used in the window size controllers 85 and 82 are different. One is 1 and the other is 0, as shown in Figs. 9A and 9β. In this way, when the bit controllers 92 and 94 transmit a preset value, the calculated similarity value will not be affected. As for how big the sliding window is when the system is started, this page depends on the experimental data. Theoretically, a sliding sight with a size of 4 yards is enough. However, in order to prevent the error caused by too much clutter, a sliding window with a larger i will have better fault tolerance. The invention has the following advantages: 1. Power saving: Because only the received copies are captured for comparison, the required doorway red AJ ^ ^ method, required for the goods C + The switching action will be relatively reduced and the name ^ is very important in automatic communication. 2. Has the ability to make mistakes. The deafness in the present invention; —-access aa Y 疋 + to find a similar sequence, not, pg-,, ^ ☆ Just in the noisy environment ancient, Tujia Missing,…,…, and 3 erroneous messages, the method of the present invention is still; ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Simultaneous: Method C boundary of the invention. Code group and time frame border.乂 Simultaneously find the base station to which the invention belongs. Although the present invention is disclosed as above with a preferred embodiment, it is not limited to

Page 17 490981

Page 18 490981 Simple illustration of the diagram »Page 19

Claims (1)

490981 VI. Scope of patent application 1. A method for code group identification and frame boundary synchronization, including the following steps: (a) Provide a secondary synchronization number of complex code length of 1 code (个) (secondary synchronization code sequences? SSCSs) 'Each SSCS corresponds to a code group; (b) Sensing and sampling the signals transmitted by a base station to form a reference code sequence (RCS) , The reference sequence contains codes with less than L samples; (c) comparing the reference sequence with the secondary synchronization code sequence to find one or more most similar sequences; (d) when the most similar sequence is not unique, increase the number of codes in the reference sequence The number of samples is returned to step (c); and (e) when the most similar sequence is unique, a code group and a frame boundary (framebund) associated with the most similar sequence are output. 2 · The method of item i in the scope of patent application, wherein the secondary synchronization code sequence is used by α δ to make its ring shifts (cyc 1 丨 c-shifts) are only 3. Item method, wherein the step (d) increases the number of samples by one each time. 4. The method as described in the first item of the patent application scope, wherein the method further comprises a step: re-execute the steps (a) to (e), and generate a final code group by majority voting The group borders with an elected box. 490981 VI. Scope of patent application 5 · The method of item 1 of the scope of patent application, wherein the method is performed on a digital signal processing chip. 6 · — Cod group identification and frame synchronization device, including: a memory to store plural code groups, the mother-code group contains a specific sequence (Predetermined code sequence); a recording device for recording a signal transmitted by a base station to form an input number (Kincoining sequence); a plurality of processors, each processor processing a corresponding code group, and each ^ one processing The device includes: a first sampling device (means for sampling) and for sampling a part of the input sequence to form a first reference sequence; a second sampling device for sampling the corresponding code of the portion Sampling the plural cyclic sequences that may be generated by the specific sequence corresponding to the group, respectively, to form a plurality of second reference sequences; and a search device that uses the first reference sequence and the second reference sequences to make a comparison To search for the most similar series of one or more single code groups; and _ a control device in the most similar series of these single code groups, Find one or more of the most similar series. When the most similar series is unique, the control device rotates out the code group and a frame phase related to the most similar series. When the most similar sequence is not unique, the control device causes the first and second sampling devices to increase their access ... On page 21, the scope of Shenyan's patent is to re-trigger these processing devices until the most similar number found by the control device is unique. Person 7. The device according to item 6 of the scope of patent application, in which the device is additionally included-a majority voting device, so that the recording device, the processors and the control device are repeatedly activated, and the relevant code group is repeatedly recorded The group and the frame edge: The number of voting methods produces an elected scrambled code group and an elected frame boundary. , Person 82 is the device according to item 6 of the patent application scope, wherein the recording device package a has a plurality of correlators. Each correlator is used to compare a plurality of positive parent codes (〇Uh〇g〇nai c) transmitted by the base station. 〇de) one of the phases with a degree of hardiness to output a similar degree signal. Contains 92 devices such as item 8 of the scope of patent applications, where the recording device includes a wise accumulator, and each box is added to accumulate a relative correlator at every other time. The degree of similarity ^ sign to output an accumulation result. Scoop person 10, such as the device of the scope of patent application item 9, wherein the frame accumulator 卞 "has a plurality of shift register modules (shj ^ t register bank), and two shift registers The module is used to memorize the The accumulation result of ^ 'and provides the accumulation result of the relative frame accumulator before the one-time frame and the relative frame accumulator. Person 11 'As the device of item 9 of the scope of patent application, where the recording device = contains a = large selector, the accumulations output by the frame accumulators ... find a maximum accumulation result, and Output a code number (c〇de number) 'to indicate a specific orthogonality corresponding to the maximum accumulation result Page 22 7. Scope of patent application Shima 〇 Baoren 12. The device according to item 6. of the patent application scope, wherein the recording device 5 has a recorder for recording the code number to form the input sequence. 13. The device according to item 6 of the scope of patent application, wherein the second sampling set generates a second reference sequence. 14. The device according to the scope of patent application, wherein the search device includes: a similarity comparator for comparing a second reference sequence with the > test sequence 'to output a similar value; and A shift register module is used to store the similar value and trigger the second sampling device to generate another second reference sequence. 15 · The device according to item 14 of the scope of patent application, wherein the similarity comparator includes: a comparator for comparing the codes at the same position in the first reference sequence and the second reference sequence one by one Whether they are the same, when the comparison result is the same, an identical signal is output; and a total line is used to accumulate the number of times that the identical signal appears. 16. · A search method for code group and time frame boundary suitable for direct-sequence CDMA communication system. The DS-CDMA system includes / base station '. It periodically sends a length of L codes. The base station sequence is the same as the dream code (SSC) sequence. The base station SSC sequence belongs to one of the X predetermined SSC sequences. The x predetermined SSC sequences are defined as X code groups. The method includes the following steps: (a ) Sensing and sampling the SSC sequence of the base station to form a length less than L Page 23 The reference sequence; ⑻ Use the reference sequence and the x pre-c sequences to make a predetermined SSC sequence, find _ or more similar sequences that are most similar to each other; cause M to take phase (C) when the most similar sequence Counties ♦ The number of samples of one yard of code, and return to step (b) 7 Shou 'added this, in the test sequence, when the most similar sequence is unique, enter a code group related to the sequence ^ and-box Boundary (fram boundary) number 17. Such as the patent " monthly range method 16 The s binary series is designed to make its cyclic shift (cycHc_shift: where step (c) is the case where the method is additionally packaged. 1) The method of item i 6 of the scope of patent application increases the number of samples each time by 1. 1 9 · The method of item 6 of the patent application includes a step: / Re-execute steps (a) to (d) to generate a final code group by μj〇rity voting Group with — (elected) box border. Take ~ 2◦ • If the patent asks for the method in the 16th range, which side Embrace a digital signal processing chip. Hold